The goal of the research presented in this report is to propose, develop and test an integrated reliability analysis to optimise the maintenance strategies of the railway industry. This integrated analysis applies traditional statistics theories as well as Bayesian statistics using Markov Chain Monte Carlo (MCMC) methodologies. Using the Bayesian inference leads to greater flexibility because such analysis can simultaneously accommodate the following:

• Small sample data;

• Incomplete data set, including censored or truncated data;

• Complex operational environments.

In this report, an integrated procedure for Bayesian reliability inference using MCMC is applied to a number of case studies using locomotive wheel degradation data from Iron Ore Line (Malmbanan), Sweden. The research explores the impact of a locomotive wheel’s installed position on its service lifetime and attempts to predict its reliability characteristics by using parametric models, non-parametric models, frailty factors, etc.